• Journal of Intelligence and Information Systems
• /
• v.9 no.1
• /
• pp.227-249
• /
• 2003

Prediction of corporate failure using past financial data is a well-documented topic. Early studies of bankruptcy prediction used statistical techniques such as multiple discriminant analysis, logit and probit. Recently, however, numerous studies have demonstrated that artificial intelligence such as neural networks can be an alternative methodology for classification problems to which traditional statistical methods have long been applied. In building neural network model, the selection of independent and dependent variables should be approached with great care and should be treated as model construction process. Irrespective of the efficiency of a teaming procedure in terms of convergence, generalization and stability, the ultimate performance of the estimator will depend on the relevance of the selected input variables and the quality of the data used. Approaches developed in statistical methods such as correlation analysis and stepwise selection method are often very useful. These methods, however, may not be the optimal ones for the development of neural network model. In this paper, we propose a genetic algorithms approach to find an optimal or near optimal input variables fur neural network modeling. The proposed approach is demonstrated by applications to bankruptcy prediction modeling. Our experimental results show that this approach increases overall classification accuracy rate significantly.

• Journal of the Korean Data and Information Science Society
• /
• v.26 no.6
• /
• pp.1353-1366
• /
• 2015

These days Shewhart control chart for evaluating stability of the process is widely used in various field. But it must follow strict assumption of distribution. In real-life problems, this assumption is often violated when many quality characteristics follow non-normal distribution. Moreover, it is more serious in multivariate quality characteristics. To overcome this problem, many researchers have studied the non-parametric control charts. Recently, SVDD (Support Vector Data Description) control chart based on RBF (Radial Basis Function) Kernel, which is called K-chart, determines description of data region on in-control process and is used in various field. But it is important to select kernel parameter or etc. in order to apply the K-chart and they must be predetermined. For this, many researchers use grid search for optimizing parameters. But it has some problems such as selecting search range, calculating cost and time, etc. In this paper, we research the efficiency of selecting parameter regions as data structure vary via simulation study and propose a new method for determining parameters so that it can be easily used and discuss a robust choice of parameters for various data structures. In addition, we apply it on the real example and evaluate its performance.

• Journal of agricultural medicine and community health
• /
• v.34 no.2
• /
• pp.256-266
• /
• 2009

Objectives: This study was to investigate the factors affecting the depression of the elderly women in poverty in community. Methods: The subjects were 1,208 elderly women over 65 years who were enrolled in the Public Health Care Center from Apr. 2008 to Jun. 2008. Data were collected using questionnaires including general characteristics, health related behaviors and health status by nurses at the time of enrollment. The collected data were analyzed by descriptive statistics, $x^2$-test, Pearson correlation coefficients and multivariate logistic regression. Results: The major findings of this study are as follows : The elderly women in poverty show a tendency to have lower level in income, education, self-rated health, cognitive function compared with ordinary women in old age. The predictors of depression of the elderly women in poverty were spouse's existence or nonexistence, type of insurance, cognitive function, and self-rated health. Conclusion: These findings suggest the need to develop nursing strategies for decreasing depression in the elderly women in poverty. To decrease the depression of the elderly women, the above-mentioned major influencing factors should be considered.

• The Korean Journal of Financial Management
• /
• v.9 no.2
• /
• pp.101-147
• /
• 1992

증권거래소(證券去來所)는 시황에 따라 위탁증거금율(委託證據金率)을 탄력적으로 변경 운용함으로써 시장의 수급을 조절하는 등의 시장관리수단의 하나로 이용하여 공정한 시세형성을 기하고자 설립시부터 증권회사로 하여금 매매의 위탁시 위탁증거금을 징수하도록 규정하고 증거금율을 상황에 따라 신축적으로 운용하여 1962년 이후에만도 무려 32회이상 변경하였다. 따라서 문제의 핵심은 위탁증거금징수가 주식시장에서의 과잉투기행위를 근절시키고 주가변동율(株價變動率)(stock volatility)을 감소시켜 공정거래질서(公正去來秩序)를 확보하는데 기여하고 있는지의 여부가 된다. 이 점은 특히 미국(美國)에서 1987년 10월 소위 '검은 월요일(Black Monday)'당시 갑작스러운 주가폭락과 시장체계의 붕괴사태이후 금융시장의 발전을 모색하는 정책당국자들과 학자들사이에 새로운 주목을 받기 시작하였다. Salinger(1989)와 Schwert(1989)는 위탁증거금율(委託證據金率)의 변경과 주가변동율(株價變動率)의 감소와는 아무런 인과관계가 없다고 결론을 내리고 있다. 특히 Schwert는 거래일시중단시책마저도 주가변동율에 별 효과가 없다고 주장하면서 금융공황과 관련된 거래일시중단은 주가변동을 큰 폭으로 증가시켜왔으나 금융공황을 동반하지 않은 기래일시중단은 높은 주가변동율과 무관함을 밝히고 있다. Hardouvelis(1991)는 그러나 위탁증거금율을 상승시키면 주가변동율이 낮아지며, 결과적으로 주가가 본원적가치(本源的價値)로부터 일탈하는 현상도 줄어든다는 사실을 통계적으로 입증하고, 위탁증거금의 징수가 시장을 교란하는 악성투기행위를 억제시키는데 매우 효과적인 정책수단이라고 주장하고 있다. 본 연구는 우리나라 주식시장에서 과잉투기현상을 억제하여 시장의 안정을 확보하는 기능으로서의 위탁증거금제도에 대해 그 경제적 효과여부를 규명하는 실증분석을 행하였다. 이 논문에서는 Schwert(1989)와 Hardouvelis(1991)의 방법을 원용하여 두가지 서로 다른 방법으로 주가변동율을 측정하여 비교하였다. 통계적 기법은 기본적으로 다변량(多變量) 회귀분석법(回歸分析法)을 택하였다. 분석의 결과로 매우 흥미로운 실증상(實證上)의 규칙성(規則性)을 발견하였다. 즉 현금시장(cash market)의 위탁증거금율이 높아지면 실제주가변동율(實際株價變動率)과 초과주가변동율(超過株價變動率)이 감소되고, 또한 유행(流行)의 경우와 마찬가지로 본원적 가치로부터의 괴리가 작아진다. 이 결과에 따르면 위탁증거금의 징수는 그 제도의 취지에 부합되고 있다. 다만 제도운용상의 이유이거나 혹은 우리나라 주식시장의 투자자들이 비합리적인 투자형태를 보임에 따라 그 정책적 효과는 때로 역기능적인 결과로 초래하였다. 그럼에도 불구하고 이 연구결과를 통하여 최소한 주식시장(株式市場)에서 위탁증거금제도는 그 제도적 의의가 여전히 있다는 사실이 확인되었다. 또한 우리나라 주식시장에서 통상 과열투기 행위가 빈번히 일어나 주식시장을 교란시킴으로써 건전한 투자풍토조성에 저해된다는 저간의 우려가 매우 커왔으나 표본 기간동안에 대하여 실증분석을 한 결과 주식시장 전체적으로 볼 때 주가변동율(株價變動率), 특히 초과주가변동율(超過株價變動率)에 미치는 영향이 그다지 심각한 정도는 아니었으며 오히려 우리나라의 주식시장은 미국시장에 비해 주가가 비교적 안정적인 수준을 유지해 왔다고 볼 수 있다.

• Clinical and Experimental Pediatrics
• /
• v.45 no.5
• /
• pp.575-587
• /
• 2002

Objective : To investigate the understanding of breast-feeding in pregnant woman and the proper way of encouraging breast-feeding. Methods : Each questionnaire included items about demographic characteristics and the understanding of breast-feeding. The questionnaires were filled up by pregnant women visiting obstetric clinics in Seoul and its vicinities, Busan, Choongjoo and Chungjoo from July 2001 to August 2001. One thousand, two hundred ninety questionnaires were analysed by Chi square tests and multiple logistic regressions. Results : The majority of pregnant women(87.4%) planned breast-feeding. Forty three percent of them had plans to breast-feed for 4-6 months. There were no differences in the level of education, the family size and the source of information about breast-feeding in planning to breast-feed (P>0.05). The main reasons for not choosing to breast-feed were returns to work(41.3%), previous failures of breast-feeding(17.4%), concerns about insufficient amount of breast milk(10.9%), breast and nipple problems(10.3%) and maternal illness(9.4%). The average score on the test of the understanding about breast-feeding was 59.7/100. The average scores on the understanding about the methods and advantages of breast-feeding were 45.3/100 and 86.1/100, respectively. The maternal status of employment, previous history of breast-feeding, the time of decision to breastfeed, person advocating breast-feeding and the understanding on the advantages of breast-feeding were significant determinant factors in planning to breast-feed(P<0.05). Conclusion : Pediatricians should take steps to make an effort to increase the breast-feeding rate and to encourage breast-feeding by timely education. Beyond the medical field, political and social supports for breast-feeding are urgently needed.

• Korean Journal of Ecology and Environment
• /
• v.33 no.4 s.92
• /
• pp.350-357
• /
• 2000

This paper evaluated the influence of point source and flow events on inorganic nitrogen fractions at 17 sites of Taechung Reservoir during 1993${\sim}$1994. Total nitrogen (TN) averaged 1.53 mg/L during the study and ranged between 0.70 and 2.56 mg/L. Dissolved inorganic nitrogen(DIN) accounted for >90% of TN regardless of season and location, indicating a nitrogen-rich system showing eutrophic${\sim}$hypereutrophic conditions. Some 67${\sim}$94% of DIN was NO$_{3}$-N, whereas mean level of NH$_{4}$-N was less than 5% of DIN. During monsoon 1993, dilution of NO$_{3}$-N was evident in the headwaters as a result of mixing of lake water with rain water, while NH$_{4}$-N increased>100% compared to the premonsoon. Values of NH$_{4}$-N had a positive correlation with rainfall (r=0.85; p<0.001) and negative correlations with theoretical water residence time(r=-0.90; p<0.001) and conductivity(r=-0.78, p<0.001), respectively. These outcomes suggest that NH$_{4}$-N came from external input from the watershed during the monsoon. In both years, mean TN was greater in the mid-lake sites than any other sites. A great amount of TN in the mid-lake was most pronounced in monsoon 1994 because of an accumulated influence of the point sources during low inflow. Overall data suggest that concentrations of TN in this system did not show large differences along the longitudinal gradients and its distributions is likely determined by point-sources rather than inflow regime.

• Journal of Hospice and Palliative Care
• /
• v.18 no.2
• /
• pp.120-127
• /
• 2015

Purpose: We aimed to investigate how serum vitamin D levels are related to survival of terminally ill cancer patients. Methods: From May 2012 through June 2013, a retrospective chart review was performed on 96 hospice patients. Univariate and multivariate logistic regression analyses were conducted to identify factors associated with severe vitamin D deficiency and Coxcy and Coxional hazard analyses were used to evaluate effects on survival. Results: The mean vitamin D level in patients was $8.60{\pm}7.16ng/ml$. Vitamin D was severely deficient (<10 ng/ml) in 75 patients (78.2%), deficient (10~20 ng/ml) in 13 patients (13.5%), relatively insufficient (21~29 ng/ml) in five patients (8.3%) and sufficient ((t ng/ml) in three patients (3.1%). Hyperbilirubinemia (${\geq}1.2g/dl$) was the only factor associated with severe vitamin D deficiency according to the multiple logistic regression analysis (Odds ratio, OR=18.48, P<0.05). Although hyperbilirubinemia showed a strong association with survival (Hazard ratio, HR=2.25, P<0.01), no association was found between severe vitamin D deficiency and survival (HR=1.15, P>0.05) in Cox's proportional hazard analysis. Conclusion: Although serum vitamin D levels were severely low in terminally ill cancer patients, we found no association between severe vitamin D deficiency and patient survival.

• Korean Journal of Clinical Pharmacy
• /
• v.22 no.2
• /
• pp.123-130
• /
• 2012

거대세포바이러스(Cytomegalovirus; CMV) 감염은 동종조혈모세포이식 환자의 주요 사망원인 중 하나이다. 용량감소전처치(Reduced-intensity conditioning; RIC)를 이용한 조혈모세포이식은 골수억제전처치(Myeloablative conditioning; MAC)에 비해 골수억제 및 면역억제가 적으므로 CMV 감염 발생율을 감소시킬 것이라 예상되었으나 예방적 면역억제요법, T세포 제거 약제의 사용 등으로 서로 상이한 결과가 보고되고 있다. 2007년 1월부터 2009년 12월까지 총 141명의 환자(MAC 113명, RIC 28명)가 동종조혈모세포이식을 받았으며, CMV 감염은 MAC 62.8%, RIC 57.1% (p = 0.310), CMV 질환은 각각 12.4%, 14.3% (p = 0.785)에서 발생하였다. CMV 감염/질환 발생빈도와 CMV 항원 혈증검사 지속기간, 초기/최고치, 생존율은 두 군간 유의한 차이가 없었다. CMV 감염 위험인자에 대한 다변량분석 결과, 환자가 고령일수록(HR 1.024, 95% CI 1.002-1.045; p = 0.031) 또는 grade 2 이상의 급성 이식편대숙주병이 발생한 경우에(HR 1.849, 95% CI 1.031-3.315; p=0.039) CMV 감염 발생 위험율이 유의하게 높았다. 결론적으로, 전처치요법 강도에 따른 CMV 감염의 발생빈도와 발현양상의 차이는 없었으나, 고령이거나 grade 2 이상의 급성 이식편대숙주병이 발생한 환자의 경우 CMV 감염 발생과 유의한 연관성을 보였다. 이상과 같은 결과에 비춰 봐서 CMV 질환이 대부분 이식 100일 이후에 발생한 점을 고려할 때, 이식 후 CMV 감염 발생 시 ganciclovir 선제요법과 함께 이들 환자들에게 지속적인 모니터링을 실시하는 것이 필요할 것으로 사료된다.

• Korean Journal of Psychosomatic Medicine
• /
• v.26 no.1
• /
• pp.68-75
• /
• 2018

Objectives : Since pain in cancer patients is an important factor that greatly affects the quality of life and prognosis, many attempts have been actively made to reduce the pain. Despite continuous effort on reducing pain after Endoscopic submucosal dissection (ESD), research has not been done on psychological factors as much as on biological factors affecting pain. The objective of this study is to investigate the psychological factors affecting postoperative pain in gastric tumor patients who underwent gastric ESD. Methods : 91 gastric tumor patients who visited National Health Service Ilsan Hospital in Korean between May 2015 and June 2016, and received ESD were evaluated. Baseline characteristics including sociodemographic factors, anxiety, depression, and resilience were evaluated before the procedure. Multivariate logistic regression was done to analyze factors affecting postoperative pain. Results : The group with high postoperative pain showed lower alcohol consumption and higher depressive symptom scores than the group with low postoperative pain. Also, the group with high postoperative pain showed lower total resilience score with lower subtotal scores in self-control and positive item. Multivariate logistic regression analysis of the postoperative pain showed that patients with lower score in self-control of resilience [odd ratio (OR), 0.911 ; 95% CI, 0.854-0.971, p=0.004) reported more pain after ESD. Conclusions : This study showed that patients with lower self-control ability of resilience felt more pain after ESD. Among the psychological factors evaluated in this study, resilience of an individual seems to have effect on pain.

• Journal of the Korean Orthopaedic Association
• /
• v.56 no.1
• /
• pp.61-67
• /
• 2021

Purpose: To evaluate the prognostic factors affecting poor functional outcomes in patients with retear after rotator cuff repair. Materials and Methods: From January 2013 to December 2018, among 631 patients who underwent arthroscopic repair of a rotator cuff tear, 42 patients, who could be followed-up for more than one year and showed a retear of the repaired cuff on magnetic resonance imaging (MRI), were collected retrospectively. The preoperative demographic data, range of motion, American Shoulder and Elbow Surgeons (ASES) score, fatty degeneration, and tear progression on postoperative MRI, as well as other factors that could affect the clinical outcomes, were analyzed. Patients who scored <80 points on the ASES score were allocated to the poor function group. The risk factors for poor clinical outcomes were compared with the group with ASES scores of 80 or above. Results: The postoperative functional results in the group with retear (n=42) after arthroscopic rotator cuff repair showed significant improvement. Univariate analysis revealed the preoperative visual analogue scale (VAS) score and tear progression to have associations with a poor shoulder function. In addition, subscapularis repair was found to be associated with a good shoulder function. The preoperative VAS score and tear progression except for subscapularis repair were independent factors associated with poor clinical outcomes according to multivariate logistic regression analysis. Conclusion: In patients with retear after rotator cuff repair, the preoperative VAS and tear progression in postoperative MRI are factors predicting a poor functional outcome.